© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 69455
    Prev
    Next
    INDEX
    Explanations

    duct tape or ductal medical conditions

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron activates on occurrences of the substring “duct” (e.g. duct, ducts, ductal, duct-tape, etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    references to ducts—both mechanical conduits and anatomical ducts—including derived forms and closely related compound terms.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    the word "duct" (or "ductal") in technical and medical contexts.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    
    -2.69
    ization
    -2.45
    錾
    -2.39
    艄
    -2.16
    ۔
    -2.06
    分で
    -2.06
    糌
    -2.06
     qualcuno
    -2.03
    킁
    -2.02
     quartier
    -1.96
    POSITIVE LOGITS
    <td>
    2.50
     Any
    2.47
    ’
    2.31
     From
    2.25
    になる
    2.00
     Before
    1.98
     Wird
    1.98
     When
    1.98
     TWO
    1.98
    家の
    1.93
    Activations Density 0.005%

    No Known Activations